Comparing big patterns and small ones

When defining your processes in xProcess you need to decide on the level of granularity at which to define task patterns and the degree of complexity to build in. Compare these two examples of feature pattern.

This first one is nice and simple. Two people will work on the feature using and/or modifying the three associated artifacts (Word documents in this case) Specification, Design and Test Spec. It's simplicity is its greatest advantage since this allows a lot of flexibility when such features are planned in a real project.

The next pattern is more complex and goes to a finer level of granularity. It defines 5 subtasks, using 4 specific role types and having different artifacts for each of the tasks. The artifacts are a mix of Word documents Wiki pages and xProcess Forms (user-defined data held in internal XML format). The result is a more complex patterns which breaks down the work into more speciifc packages and provides more guidance in terms of the work to be done.

There are dependencies defined on this pattern and there is also a gateway ("Feature complete") which defines the QA requirements for its completion.

This more detailed view of the same pattern shows the allocation of tasks to role types and the specifics of the artifacts.

Which of these two patterns is most appropriate for your projects? I can only recommend you try them and see... and most importantly let me know what you find!